Pith. sign in

REVIEW 4 major objections 5 minor 42 references

Towards Fingerprint Mosaicking Artifact Detection: A Self-Supervised Deep Learning Approach

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read This paper claims that a self-supervised segmentation model, trained only on fingerprints synthetically corrupted with shifted patches and lines, can detect hard mosaicking artifacts in fingerprint images from sensors and modalities it…

desk verdict Useful, clearly-written self-supervised detector for synthetic fingerprint mosaicking artifacts, but the central claim about real artifacts is untested because all ground truth comes from the same injection pipeline used for training. read the letter →

arxiv 2501.05034 v1 pith:LFO3UMNZ submitted 2025-01-09 cs.CV cs.LGeess.IV

classification cs.CVcs.LGeess.IV
keywords fingerprintmosaickingartifactdetectionself-supervisedlearningimagesegmentationbiometricqualityscoreout-of-distributiongeneralization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that a deep segmentation model can detect hard mosaicking artifacts in fingerprint images without any manually labeled real examples, using only self-supervised training on images the authors synthetically corrupt with shifted patches and lines. The central claim is that this suffices for the model to localize artifacts in fingerprint modalities and sensors it never saw during training, including rolled, slap, and pressed fingerprints, and to remain stable under skin defects, scars, and noise. If true, this would give biometric systems an automated quality check that catches stitching errors before they can shift minutiae and double the equal-error rate of recognition algorithms. The paper also introduces a scalar mosaicking artifact score to quantify severity.

What carries the argument

The load-bearing mechanism is the self-supervised artifact-generation pipeline, which creates input–mask pairs by selecting up to four patches (each 5–15% of the image dimensions) and offsetting them by 2–7% of the image size, or with 25% probability shifting pixels along a full vertical or horizontal line. A U-Net-style segmentation network with an encoder pretrained on a large general image dataset is trained with Jaccard (intersection-over-union) loss to reproduce these masks. The second component is the mosaicking artifact score: a weighted sum over detected patch areas and line widths, normalized by image area, whose detection threshold is the patch weight $b = 5$, meaning a score at or above 5 indicates at least one closed patch-sized misalignment.

What would settle it

Collect genuine rolled fingerprints that contain known mosaicking seams from a production acquisition system, have human experts annotate the artifact regions, and compute the model's IoU on those real masks; if the IoU falls substantially below the 0.9–0.98 range reported on synthetic artifacts, the central claim fails. A weaker but simpler check is to compare the model's artifact score distribution on real mosaicked images against expert severity ratings and see whether the threshold of 5 separates acceptable from unacceptable images.

Watch

Extended reading notes

Core claim

The discovery is that synthetic patch and line offsets, generated at training time, form a sufficient supervisory signal for a U-Net-style segmentation network to learn the concept of a hard mosaicking artifact. On the contactless test set the model reaches an IoU of 0.982, and on out-of-distribution rolled and slap fingerprints from a public benchmark it retains IoUs of 0.908 and 0.959, with accuracy near 1.000. The same model trained on only 26k pressed fingerprints performs similarly on a later pressed dataset and on rolled fingerprints from the same sensor line. The paper further reports that small stitching offsets raise the equal-error rate of three fingerprint identification systems by 112–132%, which is the practical motivation for the detector.

Load-bearing premise

The assumption that synthetic patch and line offsets, produced by the same augmentation pipeline used for training, are a faithful proxy for real mosaicking artifacts in rolled, pressed, and contactless fingerprints; this proxy is never validated against manually annotated real mosaicked images.

Editorial extensions

If this is right

  • The detector can be deployed as an automated screen for rolled fingerprints, a modality that currently lacks a dedicated quality metric, flagging images for re-acquisition before they enter central registers.
  • Because the model transfers across modalities with IoU above 0.9, a single training run on one unlabeled corpus can serve sensors of other types.
  • The mosaicking artifact score, with its threshold at the patch weight of 5, gives a portable severity measure suitable for large-scale database screening; on the pressed test set the false-match rate is 0.061%.
  • Removing or avoiding images with detected artifacts should reduce the 112–132% equal-error-rate increases that small and large offsets cause in three tested identification systems.
  • Since training needs no manual annotations, the pipeline can be retrained on new unlabeled fingerprint repositories as they become available.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The synthetic proxy leaves untested the hardest real cases, such as alpha-blended seams or elastic deformations from finger rolling; a small annotated set of real mosaicked fingerprints would reveal how much performance drops.
  • The score's weighting (patch weight 5, line weight 0.025) is a design choice; calibrating it against equal-error rate or human ratings on real data could change the threshold and improve its usefulness as a quality metric.
  • The reported false-match rate of 0.061% was measured on a test set that itself contains synthetic artifacts; the false-alarm rate on databases with genuine mosaicking errors and rare artifacts could differ.
  • A natural next step is to apply the score to historical fingerprint databases to estimate how often severe mosaicking artifacts occur in practice.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes a self-supervised deep learning method for detecting and scoring hard mosaicking artifacts in fingerprint images. A UNet++/ResNeSt-50d segmentation model is trained on large unlabeled single-shot contactless and pressed fingerprint datasets, with supervisory labels generated by a synthetic augmentation pipeline that injects patch offsets and line shifts. The authors also introduce a mosaicking artifact score S and evaluate the model on in-distribution data, on NIST 300a rolled and slap images, on pressed and rolled datasets from a later acquisition, on synthetic fingerprints with simulated skin alterations, and through an EER study of three fingerprint matchers.

Significance. If the proposed model indeed detects real mosaicking artifacts across modalities and sensors it was not trained on, this would be a practically valuable quality-control tool for rolled and contactless fingerprint acquisition, where artifacts can shift minutiae and degrade recognition accuracy. The strengths of the paper include the use of large real-world training sets, a self-supervised pipeline that avoids manual annotation, out-of-distribution evaluation on NIST 300a and subsequent acquisition batches, robustness tests against synthetic skin defects, and a multi-ABIS EER analysis. The central weakness is that every quantitative detection result is measured against ground-truth masks generated by the same synthetic artifact injection procedure that produces the training signal, so the reported IoU/F1 figures do not yet demonstrate detection of real mosaicking errors.

major comments (4)
  1. [II.B.3 and Table II] The evaluation of model performance on NIST 300a rolled and slap images, as well as on PRD-1-Test, PRD-2, and ROD-1, uses ground-truth masks created by the same patch-offset and line-shift augmentation procedure described in Section II.B.3 that generates the training labels. The IoU, F1, F2, and recall values in Table II therefore measure how well the model reproduces the synthetic artifact generator, not how well it detects real stitching seams, alpha-blended discontinuities, or non-rigid misalignments produced by actual mosaicking algorithms. This is a load-bearing point because the abstract and conclusion claim high accuracy on rolled, pressed, and contactless fingerprints. The authors should validate on independently generated realistic mosaicking errors or on expert-annotated real mosaicked images, or alternatively temper the generalization claims to synthetic artifacts only.
  2. [IV.F and V] The paper itself states that supervised finetuning on real fingerprint mosaicking artifacts and extension to real mosaicking artifact data are future work (Section IV.F, Section V). This directly acknowledges that no real ground-truth artifacts were used anywhere in the evaluation. Consequently, the robustness results in Section III.B and the score distribution analysis in Section III.A, while informative about false positives on clean images, do not establish that the model flags genuinely mis-stitched rolled or pressed fingerprints. I recommend that the manuscript either add a real-data validation study or explicitly restrict the central claims to the synthetic artifact setting.
  3. [Eq. (1) and IV.D] The mosaicking artifact score S depends on hand-chosen weights b=5 and c=0.025, and the detection threshold for real images is identified with the patch weight b. The paper validates this threshold only through synthetic experiments and through the observation that clean images have low scores; it does not calibrate the threshold against real mosaicking artifacts or against images that human experts judge as erroneous. Since the score is proposed as an automated screening tool (Section IV.D), the threshold choice should be justified with real positives and negatives, or presented as a tunable parameter with clear guidance for calibration.
  4. [III.C and Table IV] The EER analysis demonstrates that synthetic small and large offset artifacts degrade the accuracy of SourceAFIS, Bozorth3, and Innovatrics IDKit. This is a useful result, but it should not be presented as validating the detector's real-world utility: the artifacts in Table IV are generated with the same offset ratios used in training (2% to 7% for large offsets), and the table contains no comparison between images the detector flags and images that are actually contaminated by real mosaicking errors. The connection between the EER increase and the artifact score would be stronger if the EER were measured on images whose artifact scores were confirmed against independent ground truth.
minor comments (5)
  1. [Table III] The caption of Table III refers to the second model as the "CB Model", while the text and the rest of the paper call it the "PR Model"; please make the terminology consistent.
  2. [III.B] There are typos in Section III.B: "medium woulds" should be "medium wounds", and in Table III the header uses "tildelow" where a tilde symbol is intended. These should be corrected.
  3. [Eq. (1)] Equation (1) is typeset with ambiguous summation notation (e.g., "nX i") and some variables are not defined before the equation. Please rewrite the equation in standard summation form and define all symbols clearly.
  4. [References] NIST 300a is cited as reference [13] with the title "NIST special database 300"; please clarify whether the dataset is Special Database 300 or Special Publication 300a and cite the appropriate publication.
  5. [Supplementary materials] The text repeatedly refers to supplementary materials for score distributions, alteration examples, and loss plots, but the arXiv version does not appear to include them. The authors should either include the supplementary material or state where it is available.

Circularity Check

2 steps flagged · score 6.0 of 10

Real-modality evaluation is performed against the same synthetic patch/line-offset generator that defines the training labels, so the reported IoU results measure reproduction of the augmentation routine rather than detection of real mosaicking artifacts.

  1. self definitional [Section II.B.3 (Self-Supervised Learning); evaluation in Section III.A / Table II]
    "To introduce the supervisory signal, we employed two methods of adding mosaicking artifacts. The first method simulates patch based artifacts and the second one line based artifacts along the whole image width or height. The first involves selecting a patch within the image, sized between 5% and 15% of the image dimension, and offsetting its pixels by 2% to 7% of the image dimensions."

    These synthetic patch/line offsets are the only artifact labels ever generated: the same routine creates the masks used for training and the ground-truth masks used for the Table II evaluation on real-image datasets (NIST 300a rolled/slap, PRD, ROD), because no real mosaicking-artifact annotations exist. Consequently the IoU/F1/recall values in Table II measure how well the model inverts the pixel-offset generator on unseen images and modalities, not whether it detects genuine stitching seams, blending discontinuities, or non-rigid misalignments in real rolled or pressed fingerprints. The paper itself later defers 'the extension of the approach to real mosaicking artifact data' to future work, confirming that the test target is the generator's output.

  2. other [Section III.C (Effect of Mosaicking Errors on Equal-Error Rate) and Section IV.D (Mosaicking Artifact Score)]
    "The score itself is validated by EER increases observed in the presence of artifacts, as discussed in section IV-C. ... We categorized mosaicking errors into two groups based on the offset ratio of the artifacts relative to the image dimensions: 1) Small Offsets: Offsets between 1% and 2% of the input image height and width, rounded towards zero. ... 2) Large Offsets: Offsets between 2% and 7% of the input image height and width, rounded towards zero."

    The 'artifacts' whose EER impact is measured are generated by applying the same kind of rigid pixel offsets (1-7% of image dimensions) as the training augmentation in II.B.3. The EER table therefore shows matcher sensitivity to the exact synthetic transformations the model was trained to segment, not to real mosaicking failures produced by actual stitching algorithms. Using this EER result to 'validate' the proposed artifact score closes the loop back to the training-time definition of artifacts and does not independently confirm the score on real mosaicked acquisitions.

full rationale

The paper's training signal is a synthetic patch/line-displacement generator (II.B.3). All ground-truth masks used in the Table II evaluation on NIST 300a rolled/slap, PRD-1/2, and ROD-1 must come from that same generator because the paper never obtains or uses real mosaicking-artifact annotations; it explicitly defers 'supervised finetuning ... on real fingerprint mosaicking artifacts' and 'the extension of the approach to real mosaicking artifact data' to future work. The high IoU/F1 values are therefore a genuine, non-trivial result about the network's ability to invert the augmentation on unseen images and sensor modalities, but the abstract's claim that the model 'effectively identifies mosaicking errors' on rolled, pressed, and contactless fingerprints is not tested on real artifacts. The artifact-score validation via EER (III.C, IV.D) is similarly internal: the injected offsets are the same kind of transformations used in training. There is no load-bearing self-citation or imported uniqueness theorem here; the circularity is the identity of label generator and test generator, not an author-citation chain. Because the segmentation task itself has independent measurable content (the network could fail on new images), this is partial circularity rather than a fully forced derivation, so a score of 6 is appropriate.

Assumptions & free parameters 5 free parameters · 4 assumptions · 1 invented entities

The central claim rests on the assumption that synthetic patch and line offsets reproduce real mosaicking artifacts; the score uses hand-chosen weights b=5 and c=0.025; training backgrounds are assumed artifact-free. No physical entities are invented; the artifact score is a defined metric without independent external validation.

free parameters (5)
  • Patch weight b = 5
    Hand-chosen in Eq. (1) to weight patch artifacts; it also defines the detection threshold, so the 'true artifact' criterion is an input, not an output.
  • Line artifact weight c = 0.025
    Hand-chosen in Eq. (1) to balance line artifacts against patch artifacts in the score.
  • Patch size range = 5-15% of image dimension
    Chosen for the synthetic artifact generator in II.B.3; determines what the model learns to detect.
  • Patch offset range = 2-7% of image dimensions
    Chosen for the synthetic artifact generator; the same range is used in the EER experiment for 'large offsets', coupling the impact study to the training distribution.
  • Line artifact probability = 25%
    Chosen in II.B.3 to balance line vs patch artifacts in the training signal.
assumptions (4)
  • domain assumption Single-shot contactless fingerprint images and pressed FTIR images are free of mosaicking artifacts, so they can serve as clean backgrounds for synthetic artifact injection.
    Stated in II.A: the training data are single-shot images 'without mosaicking artifacts' and 'we assume that no mosaicking artifacts are introduced while capturing pressed fingerprints'. If this is false, the training signal is contaminated.
  • domain assumption The patch and line offset injection procedure produces artifacts that are representative of real hard mosaicking errors.
    This is the load-bearing proxy assumption in II.B.3; it is never checked against real mosaicked fingerprints.
  • ad hoc to paper The hand-chosen score weights b and c (and the threshold b=5) provide a meaningful severity measure for real artifacts.
    Defined in II.C and IV.D; the score is a proposed metric, not derived, and its real-world validity is assumed when interpreting score bands on real data.
  • domain assumption NIST 300a rolled and slap images, after synthetic artifact injection, are a valid out-of-distribution test for real rolling mosaicking errors.
    Used in III.A; the test measures detection of synthetic distortions on scanned ink prints, not real stitching discontinuities from a mosaicking algorithm.
invented entities (1)
  • Mosaicking artifact score S
    purpose: Quantifies the severity of detected artifacts from the segmentation mask, for automated quality control.
    Defined in Eq. (1) with hand-chosen weights b=5 and c=0.025; the paper validates it via EER increases on the same synthetic artifacts, not on independently annotated real mosaicking errors, so its external validity is unestablished.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Fingerprint Mosaicking Artifact Detection: A Self-Supervised Deep Learning Approach." pith.science (2026). https://pith.science/paper/LFO3UMNZ

@misc{pith2026250105034,
  author       = {Pith},
  title        = {Pith review of: Towards Fingerprint Mosaicking Artifact Detection: A Self-Supervised Deep Learning Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LFO3UMNZ}},
  note         = {Machine review of arXiv:2501.05034}
}
read the original abstract

Fingerprint mosaicking, which is the process of combining multiple fingerprint images into a single master fingerprint, is an essential process in modern biometric systems. However, it is prone to errors that can significantly degrade fingerprint image quality. This paper proposes a novel deep learning-based approach to detect and score mosaicking artifacts in fingerprint images. Our method leverages a self-supervised learning framework to train a model on large-scale unlabeled fingerprint data, eliminating the need for manual artifact annotation. The proposed model effectively identifies mosaicking errors, achieving high accuracy on various fingerprint modalities, including contactless, rolled, and pressed fingerprints and furthermore proves to be robust to different data sources. Additionally, we introduce a novel mosaicking artifact score to quantify the severity of errors, enabling automated evaluation of fingerprint images. By addressing the challenges of mosaicking artifact detection, our work contributes to improving the accuracy and reliability of fingerprint-based biometric systems.

Figures

Figures reproduced from arXiv: 2501.05034 by the authors.

Figure 1
Figure 1. Depiction of different artifact types created for super [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Model architecture of our proposed combination of [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Exemplary results with blurred input images for privacy [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Synthetic fingerprint alterations based on SynColFinGe [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 38 canonical work pages

  1. [1]

    Image Stitching based on Feature Extraction Techniques: A Survey

    Ebtsam Adel, Mohammed Elmogy, and Hazem Elbakry. “Image Stitching based on Feature Extraction Techniques: A Survey”. In: International Journal of Computer Applications 99.6 (Aug. 20, 2014), pp. 1–8. ISSN : 09758887. DOI: 10 . 5120 / 17374 - 7818. URL: http : //research.ijcaonline.org/volume99/number6/pxc3897818.pdf (visited on 11/27/2024)

  2. [2]

    Overview of fingerprint recognition system

    Mouad. M.H. Ali et al. “Overview of fingerprint recognition system”. In: 2016 International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT). 2016 International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT). Mar. 2016, pp. 1334–1338. DOI: 10.1109/ICEEOT.2016.7754900. URL: https://ieeexplore.ieee.org...

  3. [3]

    Objective image quality assessment of 3D synthesized views

    Federica Battisti et al. “Objective image quality assessment of 3D synthesized views”. In: Signal Processing: Image Communication 30 (Jan. 31, 2015), pp. 78–88. DOI: 10.1016/j.image.2014.10.005

  4. [4]

    Novel algorithm for finger- print mosaicing using phase correlation method

    Satish H. Bhati and Umesh C. Pati. “Novel algorithm for finger- print mosaicing using phase correlation method”. In: 2015 Global Conference on Communication Technologies (GCCT) . 2015 Global Conference on Communication Technologies (GCCT). Apr. 2015, pp. 29–33. DOI: 10 . 1109 / GCCT . 2015 . 7342618. URL: https : / / ieeexplore.ieee.org/document/7342618 (...

  5. [5]

    Synthetic Fingerprint Generation

    Raffaele Capelli. “Synthetic Fingerprint Generation”. In: Davide Mal- toni et al. Handbook of Fingerprint Recognition . London: Springer London, 2009, pp. 271–302. ISBN : 978-1-84882-253-5 978-1-84882- 254-2. DOI: 10.1007/978-1-84882-254-2 6. URL: http://link.springer. com/10.1007/978-1-84882-254-2 6 (visited on 10/02/2023). 10

  6. [6]

    Successive minutia-free mosaicking for small- sized fingerprint recognition

    Jiong Chen et al. “Successive minutia-free mosaicking for small- sized fingerprint recognition”. In: IET Image Processing 13.7 (2019). eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1049/iet- ipr.2018.5972, pp. 1146–1151. ISSN : 1751-9667. DOI: 10.1049/iet- ipr.2018.5972. URL: https://onlinelibrary.wiley.com/doi/abs/10.1049/ iet-ipr.2018.5972 (visited...

  7. [7]

    Fingerprint Mosaicking by Rolling and Sliding

    Kyoungtaek Choi, Hee-seung Choi, and Jaihie Kim. “Fingerprint Mosaicking by Rolling and Sliding”. In: Audio- and Video-Based Biometric Person Authentication . Ed. by Takeo Kanade, Anil Jain, and Nalini K. Ratha. Berlin, Heidelberg: Springer, 2005, pp. 260–269. ISBN : 978-3-540-31638-1. DOI: 10.1007/11527923 27

  8. [8]

    More than Zero: Accounting for Error in Latent Fingerprint Identification

    Simon A. Cole. “More than Zero: Accounting for Error in Latent Fingerprint Identification”. In: The Journal of Criminal Law and Criminology (1973-) 95.3 (2005). Publisher: Northwestern University School of Law, pp. 985–1078. ISSN : 00914169. URL: http://www. jstor.org/stable/3491332 (visited on 09/04/2024)

Show all 42 references
  1. [9]

    Objective view synthesis quality assessment

    Pierre-Henri Conze, Philippe Robert, and Luce Morin. “Objective view synthesis quality assessment”. In: Proceedings of SPIE - The International Society for Optical Engineering 8288 (Feb. 1, 2012), p. 53. DOI: 10.1117/12.908762

  2. [10]

    Dense Registration and Mosaicking of Fingerprints by Training an End-to-End Network

    Zhe Cui, Jianjiang Feng, and Jie Zhou. “Dense Registration and Mosaicking of Fingerprints by Training an End-to-End Network”. In: IEEE Transactions on Information Forensics and Security 16 (2021), pp. 627–642. ISSN : 1556-6013, 1556-6021. DOI: 10.1109/TIFS.2020. 3017926. arXiv...

  3. [11]

    Invisible Stitch: Generating Smooth 3D Scenes with Depth Inpainting

    Paul Engstler et al. Invisible Stitch: Generating Smooth 3D Scenes with Depth Inpainting . eprint: 2404.19758. 2024. URL: https://arxiv. org/abs/2404.19758

  4. [12]

    A Mosaic Approach to Touchless Fingerprint Image with Multiple Views

    Han Fengling, Mohammed Alkhathami, and Ron van Schyndel. A Mosaic Approach to Touchless Fingerprint Image with Multiple Views. Dec. 1, 2014. DOI: 10.1145/2659021.2659065

  5. [13]

    NIST special database 300: uncompressed plain and rolled images from fingerprint cards

    Gregory Fiumara et al. NIST special database 300: uncompressed plain and rolled images from fingerprint cards . NIST TN 1993. Gaithersburg, MD: National Institute of Standards and Technology, June 2018, NIST TN 1993. DOI: 10 . 6028 / NIST. TN . 1993. URL: http://nvlpubs.nist.g...

  6. [14]

    Segmentation Models Pytorch

    Pavel Iakubovskii. Segmentation Models Pytorch . 2019. URL: https: //github.com/qubvel/segmentation models.pytorch

  7. [15]

    Innovatrics

    IDKit SDK. Innovatrics. URL: https://support.innovatrics.com/support/ solutions/folders/5000267908 (visited on 03/06/2023)

  8. [16]

    Fingerprint mosaicking

    Anil Jain and Arun Ross. “Fingerprint mosaicking”. In: 2002 IEEE International Conference on Acoustics, Speech, and Signal Process- ing. 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing. V ol. 4. ISSN: 1520-6149. May 2002, pp. IV– 4064–IV–4067. DO...

  9. [17]

    Guidance for evaluating contactless fingerprint acquisition devices

    John Libert et al. Guidance for evaluating contactless fingerprint acquisition devices . NIST SP 500-305. Gaithersburg, MD: National Institute of Standards and Technology, July 2018, NIST SP 500–305. DOI: 10 . 6028 / NIST. SP. 500 - 305.URL: https : / / nvlpubs . nist . gov / ...

  10. [18]

    A 1D spectral image validationverification metric for fingerprints

    John M Libert, John Grantham, and Shahram Orandi. A 1D spectral image validationverification metric for fingerprints . NIST IR 7599. Edition: 0. Gaithersburg, MD: National Institute of Standards and Technology, 2009, NIST IR 7599. DOI: 10.6028/NIST.IR.7599. URL: https://nvlpub...

  11. [19]

    Touchless Multiview Fingerprint Acquisition and Mosaicking

    Feng Liu et al. “Touchless Multiview Fingerprint Acquisition and Mosaicking”. In: IEEE Transactions on Instrumentation and Mea- surement 62.9 (Sept. 2013). Conference Name: IEEE Transactions on Instrumentation and Measurement, pp. 2492–2502. ISSN : 1557-

  12. [20]

    The permanence of friction ridge skin and persistence of friction ridge skin and impressions: A comprehensive review and new results

    Keith L. Monson et al. “The permanence of friction ridge skin and persistence of friction ridge skin and impressions: A comprehensive review and new results”. In: Forensic Science International 297 (Apr. 1, 2019), pp. 111–131. ISSN : 0379-0738. DOI: 10 . 1016 / j . forsciint.2...

  13. [21]

    Error De- tection in Panoramic Videos: a Pairwise Assessment within Stitching

    Sandra Nabil, Fr ´ed´eric Devernay, and James L. Crowley. “Error De- tection in Panoramic Videos: a Pairwise Assessment within Stitching”

  14. [22]

    Proprietary Fingerprint Template (PFT) III

    NIST. Proprietary Fingerprint Template (PFT) III . Proprietary Fin- gerprint Template (PFT) III. Last Modified: 2022-08-26T14:17-04:00. Mar. 7, 2019. URL: https : / / www. nist . gov / itl / iad / image - group / proprietary-fingerprint-template-pft-iii (visited on 11/06/2024)

  15. [23]

    Optimal Decisions from Probabilistic Models: The Intersection-over-Union Case

    Sebastian Nowozin. “Optimal Decisions from Probabilistic Models: The Intersection-over-Union Case”. In: 2014 IEEE Conference on Computer Vision and Pattern Recognition. 2014 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Columbus, OH, USA: IEEE, June 2014, ...

  16. [24]

    MCLFIQ: Mobile Contactless Fingerprint Image Quality

    Jannis Priesnitz et al. MCLFIQ: Mobile Contactless Fingerprint Image Quality . Dec. 12, 2023. DOI: 10 . 48550 / arXiv. 2304 . 14123. arXiv: 2304.14123[cs]. URL: http://arxiv.org/abs/2304.14123 (visited on 02/28/2024)

  17. [25]

    SynCoLFinGer: Synthetic contactless finger- print generator

    Jannis Priesnitz et al. “SynCoLFinGer: Synthetic contactless finger- print generator”. In: Pattern Recognition Letters 157 (May 1, 2022), pp. 127–134. ISSN : 0167-8655. DOI: 10 . 1016 / j . patrec . 2022 . 04

  18. [26]

    U-Net: Convo- lutional Networks for Biomedical Image Segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. “U-Net: Convo- lutional Networks for Biomedical Image Segmentation”. In: Medical Image Computing and Computer-Assisted Intervention – MICCAI

  19. [27]

    Image versus feature mosaicing: A case study in fingerprints

    Arun Ross, Samir Shah, and Jidnya Shah. “Image versus feature mosaicing: A case study in fingerprints”. In: Proceedings of SPIE - The International Society for Optical Engineering 6202 (Apr. 17, 2006). DOI: 10.1117/12.666278

  20. [28]

    sciencedirect

    URL: https : / / www . sciencedirect . com / science / article / pii / S0167865522000915 (visited on 01/08/2024)

  21. [29]

    Improving Sensor Interoperability between Contactless and Contact-Based Fingerprints Using Pose Correction and Unwarping

    Laurenz Ruzicka et al. “Improving Sensor Interoperability between Contactless and Contact-Based Fingerprints Using Pose Correction and Unwarping”. In: IET Biometrics 2023 (Dec. 18, 2023). Ed. by Vincenzo Conti. Publisher: Hindawi, p. 7519499. ISSN : 2047-4938. DOI: 10.1049/202...

  22. [30]

    Development of Automated Image Stitching System for Radiographic Images

    Salbiah Samsudin et al. “Development of Automated Image Stitching System for Radiographic Images”. In: Journal of Digital Imaging 26.2 (Apr. 2013), pp. 361–370. ISSN : 0897-1889. DOI: 10.1007/s10278- 012 - 9483 - 5. URL: https : / / www. ncbi . nlm . nih . gov / pmc / articles...

  23. [31]

    FingerNet: An unified deep network for fingerprint minutiae extraction

    Yao Tang et al. “FingerNet: An unified deep network for fingerprint minutiae extraction”. In: 2017 IEEE International Joint Conference on Biometrics (IJCB) . 2017 IEEE International Joint Conference on Biometrics (IJCB). Denver, CO: IEEE, Oct. 2017, pp. 108–116. ISBN : 978-1-5...

  24. [32]

    Fingerprint Processing: Running the NIST NBIS Fingerprint Toolset on an ODROID-XU4

    Andrew Ruggeri. Fingerprint Processing: Running the NIST NBIS Fingerprint Toolset on an ODROID-XU4 . ODROID Magazine. Sec- tion: ODROID-XU4. May 1, 2020. URL: https://magazine.odroid. com/article/fingerprint-processing-running-the-nist-nbis-fingerprint- toolset-on-an-odroid-xu...

  25. [33]

    S-3: A Spectral and Spatial Measure of Local Perceived Sharpness in Natural Images

    Cuong Vu, Thien Phan, and Damon Chandler. “S-3: A Spectral and Spatial Measure of Local Perceived Sharpness in Natural Images”. In: Image Processing, IEEE Transactions on 21 (Apr. 1, 2012), pp. 934–

  26. [34]

    Structural Similarity Based Image Quality Assessment

    Zhou Wang, Alan Bovik, and Hamid Sheikh. “Structural Similarity Based Image Quality Assessment”. In: Digital Video Image Quality and Perceptual Coding, Ser. Series in Signal Processing and Commu- nications (Nov. 18, 2005). ISSN : 978-0-8247-2777-2. DOI: 10.1201/ 9781420027822.ch7

  27. [35]

    User’s guide to NIST biometric image software (NBIS)

    Craig I Watson et al. User’s guide to NIST biometric image software (NBIS). NIST IR 7392. Edition: 0. Gaithersburg, MD: National Institute of Standards and Technology, 2007, NIST IR 7392. DOI: 10 . 6028 / NIST. IR . 7392.URL: https : / / nvlpubs . nist . gov / nistpubs / Legac...

  28. [36]

    SourceAFIS fingerprint matcher

    Vazan. SourceAFIS fingerprint matcher . URL: https : / / sourceafis . machinezoo.com/ (visited on 01/22/2024)

  29. [37]

    Subjective and objective quality assessment of panoramic videos in virtual reality environments

    Bo Zhang et al. “Subjective and objective quality assessment of panoramic videos in virtual reality environments”. In: 2017 IEEE In- ternational Conference on Multimedia & Expo Workshops (ICMEW) . 11 2017 IEEE International Conference on Multimedia & Expo Work- shops (ICMEW). ...

  30. [41]

    Case study of the acquisition of contactless fingerprints in a real police setting

    Axel Weissenfeld et al. “Case study of the acquisition of contactless fingerprints in a real police setting”. In: 2022 International Con- ference of the Biometrics Special Interest Group (BIOSIG) . 2022 International Conference of the Biometrics Special Interest Group (BIOSIG)...

  31. [945]

    DOI: 10.1109/TIP.2011.2169974

  32. [2015]

    by Nassir Navab et al

    Ed. by Nassir Navab et al. Lecture Notes in Computer Science. Cham: Springer International Publishing, 2015, pp. 234–241. ISBN : 978-3-319-24574-4. DOI: 10.1007/978-3-319-24574-4 28

  33. [2017]

    URL: https://hal.science/hal-01849267 (visited on 07/11/2024)

  34. [9662]

    URL: https://ieeexplore.ieee

    DOI: 10.1109/TIM.2013.2258248. URL: https://ieeexplore.ieee. org/document/6562785 (visited on 07/11/2024)

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.